Also known as: ML in tariff classification, Automated classification
A branch of artificial intelligence that trains statistical models on historical tariff classification data to automatically predict the correct tariff code for goods based on their description, composition, use, and technical characteristics.
Machine learning applied to tariff classification uses algorithms that learn from millions of prior classifications—historical customs entries, SAT rulings, classification opinions, Harmonized System explanatory notes—to predict the most likely tariff code for a product. Unlike keyword searches, these models understand context, synonyms, and semantic relationships between merchandise descriptions.
TariffPro by Camtom uses machine learning models trained on over 40,000 real operations, achieving over 95% accuracy in tariff classification for the Harmonized System and TIGIE.
Machine learning does not replace the expert human classifier but empowers them: it reduces initial research time, detects potential errors, and ensures consistency in bulk classifications. Human oversight remains essential for new, complex, or ambiguously classified merchandise.
Classify with TariffPro
Learn more→Tariff Classification
The process of assigning the correct tariff code to a product within a country's tariff schedule, determining applicable duties and regulations.
TechnologyCustoms AI
The application of artificial intelligence algorithms—machine learning, natural language processing, and computer vision—to automate and optimize foreign trade processes such as tariff classification, document review, and customs risk detection.
TariffsTariff Classification Code
An 8 or 10-digit numerical code that identifies a specific product within Mexico's tariff schedule (TIGIE), determining applicable duties and regulations.
TariffsHarmonized System (HS)
An international 6-digit nomenclature developed by the WCO for classifying goods in international trade, used by over 200 countries worldwide.